Reads architectural PDFs, detects the scale on its own, measures only the trade you work in, and returns quantities, labour hours and a bid — then pushes the work into your CRM.
Takeoffs that took weeks come back in minutes — with every measurement shown, checkable, and correctable in about a minute a drawing.
Before a job can be priced it has to be counted — fixtures, runs and square footage, sheet by sheet. Four things kept that slow, and kept automation from fixing it.
Takeoff is measured in hours and days. An estimator opens a drawing set, sets the scale by hand, and works page by page counting fixtures, tracing linear runs and totalling areas — then repeats the whole thing when the set is revised.
Most of a drawing set is irrelevant to any one trade. An electrical estimator does not need the landscaping sheets, but still has to open them to find out. The volume of pages is itself part of the cost.
And generic automation does not fit. Exteriors, plumbing, electrical and mechanical each count different things in different ways, so a tool that measures everything the same way produces numbers nobody can bid from.
And an estimator has to sign their name to the bid. Every competing tool returns a number in a table with nothing behind it, which is precisely why estimators do not use them — so the requirement was never accuracy alone, it was showing the measurement on the drawing and making it correctable in seconds.
Scale is read from the drawing automatically rather than typed in, with the estimator able to override it. Getting this wrong invalidates every measurement on the sheet, so it is the first thing the system establishes and the first thing it lets you correct.
The system classifies the drawing set and keeps only the sheets relevant to the trade being estimated, so an electrical takeoff is not carried out across architectural and landscape pages.
Square footages, linear footages and material counts are extracted according to the workflow of the discipline — exteriors, plumbing, electrical or mechanical — rather than one generic measurement pass applied to everything.
Each trade counts different things in different ways. One generic pass over everything produces numbers nobody can bid from.
Floor area masked, conduit runs traced, devices counted.
Every drawing comes back with the system’s own reading painted onto it — a mask over exactly the area it attributed to each material, node by node. You are never asked to trust a number you cannot see: the measurement sits on the drawing itself, inspectable, before anything downstream depends on it. That is the part no competing tool offers.
| Item | Model | Now |
|---|---|---|
| Room 101 mask | 3,496 SF | 3,496 SF |
| Area · sheet | 6,796 SF | 6,796 SF |
| Conduit run | 194 LF | 194 LF |
| Devices | 12 no. | 12 no. |
| Labour hours | 108 h | 108 h |
| Calculated bid | 88,230 | 88,230 |
Nothing is retyped. The estimator moves the boundary; areas, labour hours and the bid follow it. Devices and runs stay put because the correction did not touch them.
Where the mask falls short, the estimator drags its nodes to cover the true extent. It is a direct manipulation on the drawing rather than a form to fill or a value to retype, which is why a drawing that needs fixing takes about a minute rather than being sent back for another pass.
Adjusting a boundary does not just change one figure. Areas, linear runs, material counts, labour hours and the bid all recompute from the corrected geometry automatically — so the estimator fixes the measurement and nothing else, and no total is left carrying the old number.
Every measurement is shown on the drawing and correctable in about a minute — so the estimator signs their name to a number they have seen, not one they were handed.
Quantities resolve into estimated labour hours and a calculated bid, then push into the CRM — assigning tasks to the relevant departments or people so the estimate becomes a job rather than a document.
| Item | Qty | Unit | Labour |
|---|---|---|---|
| Area | 7,480 | SF | 94 h |
| Conduit run | 912 | LF | 78 h |
| Devices | 148 | no. | 74 h |
| Labour hours | 246 h | ||
Until it is pushed, the estimate is a document on someone's desk.
The full scope, grouped by the stage of the estimate it serves — from reading the drawing set, through measurement and estimator review, to the job landing in your CRM.
Everything before a measurement is taken — and the two places a takeoff silently goes wrong if nobody checks them.
Each trade counts different things in different ways. One generic pass over everything produces numbers nobody can bid from.
Every measurement is shown on the drawing and correctable in about a minute — so the estimator signs their name to a number they have seen, not one they were handed.
Quantities are not the deliverable. A priced bid and assigned work are.
A drawing set is ingested and classified, irrelevant sheets are dropped, scale is detected per sheet, and the trade-specific extraction pass measures areas, lengths and counts. Estimator corrections feed back into recalibration before quantities resolve into labour hours, a bid, and CRM tasks.
Architectural PDF ingestion with automatic scale detection and manual override.
Trade-aware page filtering that discards sheets irrelevant to the discipline being estimated.
Configurable workflows for exteriors, plumbing, electrical and mechanical.
Square footage, linear footage and material count extraction.
Per-material mask overlays on every drawing, showing exactly the area the model attributed to each trade.
Node-level editing directly on the drawing — drag the boundary to the true extent, no retyping.
Around a minute to correct a drawing that needs it.
Automatic recalculation of every downstream figure the correction affects: areas, runs, counts, labour hours and the bid.
Estimated labour hours derived from the measured quantities.
Automatic bid calculation.
CRM integration with task assignment to departments and individuals.
Output formatted to match the deliverable estimators already produce by hand.
Takeoffs returned in minutes rather than weeks
Scale established automatically instead of set by hand
Only the sheets relevant to the trade are processed
Every measurement visible as a mask on the drawing, not a number in a table
Corrections made by dragging a boundary, with every affected total recomputed automatically
An estimator stays accountable for the number, which is the only way the output is signable
Estimates converted into assigned CRM tasks rather than a static document
Three modules that read a feasibility study, judge it, and write a new one from scratch
Three modules: one reads a feasibility and fills the whole Excel model, one benchmarks it against world standards and returns a gap analysis, and one generates an entirely new professional feasibility from historical data.
A walkthrough video in, a removals quote in seconds
The customer films their own home. Computer vision identifies and counts every object, estimates dismantling, handling and transport time, and returns a priced quote.
It calls every customer, then tells the board what they said
An AI agent calls every customer after service and gathers the review on the call itself. Every conversation is then analysed automatically and clustered into a picture of what is happening per department, per individual and across the whole organisation.